A video tracking method comprises: determining a first tracking target in a video, and performing a video tracking process on the first tracking target based on the video; performing area overlap degree calculation on the first tracking target in a current image and the first tracking target in a past image during the video tracking process, the past image being an image obtained when the first tracking target was tracked last time in the video tracking process; extracting respective feature information of feature points of the first tracking target in the video during the video tracking process on the first tracking target; matching the extracted respective feature information of the feature points of the first tracking target in the current image with the respective feature information of corresponding feature points of the first tracking target in the past image if a result of the area overlap degree calculation is less than or equal to a preset threshold; and continuing the video tracking process on the first tracking target if the respective feature information of at least a preset number of feature points among the extracted feature points in the current image matches with the feature information of corresponding feature points of the first tracking target in the past image.
Legal claims defining the scope of protection, as filed with the USPTO.
1. A video tracking method, comprising: determining a first tracking target in a video, and performing a video tracking process on the first tracking target based on the video; performing area overlap amount calculation on the first tracking target in a current image and the first tracking target in a past image during the video tracking process, the past image being an image obtained when the first tracking target was tracked last time in the video tracking process; extracting respective feature information of feature points of the first tracking target in the video during the video tracking process on the first tracking target; if a result of the area overlap amount calculation is less than or equal to a preset threshold, matching the extracted respective feature information of the feature points of the first tracking target in the current image with the respective feature information of corresponding feature points of the first tracking target in the past image, if the extracted respective feature information of at least a preset number of feature points among the feature points in the current image matches with the feature information of corresponding feature points of the first tracking target in the past image, continuing the video tracking process on the first tracking target.
2. The video tracking method of claim 1 , wherein extracting respective feature information of feature points of the first tracking target in the video comprises: extracting the respective feature information of the feature points of the first tracking target in the current image of the video if a result of the area overlap amount calculation of the current image of the video is less than or equal to the preset threshold; or, extracting the respective feature information of the feature points of the first tracking target in each frame image of the video in real time.
3. The video tracking method of claim 1 , wherein extracting the respective feature information of the feature points of the first tracking target in the video comprises: recognizing the feature points of the first tracking target in the video; and extracting the respective feature information of the recognized feature points.
4. The video tracking method of claim 1 , further comprising: determining the first tracking target is not a first tracking object if a number of matched feature points among the feature points in the current image is less than a preset number.
5. The video tracking method of claim 1 , further comprising: creating a second tracking target based on the feature points in the current image, and performing video tracking on the second tracking target based on the video if a number of matched feature points among the feature points in the current image is less than a preset number.
6. The video tracking method of claim 1 , after performing the area overlap amount calculation on the first tracking target in the current image and the first tracking target in the past image, further comprising: if the result of the area overlap amount calculation is greater than the preset threshold, continuing the video tracking process on the first tracking target.
7. The video tracking method of claim 1 , wherein the first tracking target is a human face object, and the feature points comprise one or more of the following items: a left eye feature point, a right eye feature point, a nose tip feature point, a left mouth corner feature point and a right mouth corner feature point.
8. The video tracking method of claim 1 , wherein the respective feature information of the feature points comprises one or more of the following items: SIFT feature information, FAST feature information and SURF feature information.
9. An object recognition method, comprising: determining a first tracking target in a video, and performing a video tracking process on the first tracking target based on the video; performing area overlap amount calculation on the first tracking target in a current image and the first tracking target in a past image during the video tracking process, the past image being an image obtained when the first tracking target was tracked last time in the video tracking process; extracting respective feature information of feature points of the first tracking target in the video during the video tracking process on the first tracking target; matching the extracted respective feature information of the feature points of the first tracking target in the current image with the respective feature information of corresponding feature points of the first tracking target in the past image if a result of the area overlap amount calculation is less than or equal to a preset threshold; and continuing the video tracking process on the first tracking target if the extracted respective feature information of at least a preset number of feature points among the feature points in the current image matches with the feature information of corresponding feature points of the first tracking target in the past image; further comprising: extracting object recognition feature vectors based on at least one frame image of the video on the first tracking target; matching the extracted object recognition feature vectors with object feature vectors of registered objects in a registration object library, the registration object library being established based on pre-learning of the object feature vectors in images of the registered objects; and taking one of the registered objects as a recognition result if the extracted object recognition feature vectors match with the object feature vectors of the one of the registered objects.
10. The object recognition method of claim 9 , before matching the extracted object recognition feature vectors with the object feature vectors of the registered objects in the registration object library, further comprising: extracting attribute information based on at least one frame image of the video on the first tracking target; filtering out registered objects corresponding to the extracted attribute information in the registration object library, the registration object library being established based on pre-learning of the object feature vectors and attribute information in images of the registered objects; wherein matching the extracted object recognition feature vectors with object feature vectors of registered objects in a registration object library comprises: matching the extracted object recognition feature vectors with respective object feature vectors of the filtered registered objects.
11. The object recognition method of claim 10 , wherein the first tracking target is a human face object, and the attribute information comprises one or more of gender information and age information.
12. The object recognition method of claim 9 , wherein the at least one frame image used for extracting the object recognition feature vectors is picked from a photo stream formed by performing the video tracking process on the first tracking target, picking conditions comprising one or more of the following items: a degree of blurring, a degree of exposure and object posture.
13. The object recognition method of claim 12 , after extracting the respective feature information of the feature points of the first tracking target in the video, further comprising: aligning images of the video with a preset image template based on positions of the feature points in the images of the video; wherein the at least one frame image used for extracting the object recognition feature vectors is picked from the images of the video which have been aligned.
14. A video tracking device, comprising: a processor; and a memory on which a computer program instruction is stored, wherein when the computer program instruction is executed by the processor, the processor performs the following steps: determining a first tracking target in a video, and performing a video tracking process on the first tracking target based on the video; performing area overlap amount calculation on the first tracking target in a current image and the first tracking target in a past image during the video tracking process, the past image being an image obtained when the first tracking target was tracked last time in the video tracking process; extracting respective feature information of feature points of the first tracking target in the video during the video tracking process on the first tracking target; if a result of the area overlap amount calculation is less than or equal to a preset threshold, matching the extracted respective feature information of the feature points of the first tracking target in the current image with the respective feature information of corresponding feature points of the first tracking target in the past image; and if the extracted respective feature information of at least a preset number of feature points among the feature points in the current image matches with the feature information of corresponding feature points of the first tracking target in the past image, continuing the video tracking process on the first tracking target.
15. The video tracking device of claim 14 , wherein extracting the respective feature information of the feature points of the first tracking target in the video comprises: extracting the respective feature information of the feature points of the first tracking target in the current image of the video if a result of the area overlap amount calculation of the current image of the video is less than or equal to the preset threshold; or, extracting the respective feature information of the feature points of the first tracking target in each frame image of the video in real time.
16. The video tracking device of claim 14 , wherein extracting the respective feature information of the feature points of the first tracking target in the video comprises: recognizing the feature points of the first tracking target in the video; and extracting the respective feature information of the recognized feature points.
17. The video tracking device of claim 14 , wherein the processor further performs the following step: determining the first tracking target is not a first tracking object if a number of matched feature points among the feature points in the current image is less than a preset number.
18. The video tracking device of claim 14 , wherein the processor further performs the following steps: creating a second tracking target based on the feature points in the current image, and performing video tracking on the second tracking target based on the video if a number of matched feature points among the feature points in the current image is less than a preset number.
19. The video tracking device of claim 14 , after performing the area overlap amount calculation on the first tracking target in the current image and the first tracking target in the past image, the processor further performing the following step: if the result of the area overlap amount calculation is greater than the preset threshold, directly continuing the video tracking process on the first tracking target.
20. The video tracking device of claim 14 , wherein the first tracking target is a human face object, the feature points comprise one or more of the following items: a left eye feature point, a right eye feature point, a nose tip feature point, a left mouth corner feature point and a right mouth corner feature point, and the respective feature information of the feature points comprises one or more of the following items: SIFT feature information, FAST feature information and SURF feature information.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
March 1, 2018
July 28, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.